doi:10.1016/j.compbiolchem.2005.08.003
Copyright © 2005 Elsevier Ltd All rights reserved.
A family of metrics for biopolymers based on counting independent sets
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
J. Rocha
,
, M. Llabrés
and R. Alberich
Department of Mathematics and Computer Science, Research Institute of Health Science (IUNICS), University of the Balearic Islands, 07122 Palma de Mallorca, Spain
Received 12 May 2005;
revised 9 August 2005;
accepted 9 August 2005.
Available online 10 October 2005.
Abstract
We introduce a new family of metrics for graphs of fixed size, based on counting-independent sets. Our definition is simpler and easier to calculate than the edge ideal metric family defined by Llabrés and Rosselló without loosing any of its abstract properties. We contrast them on some examples with graphs that represent protein secondary and three-dimensional (3D) structures. We conclude that although the edge ideal metrics are faster to calculate on some sparse graphs, in general, the independent set metrics are more tractable.
Keywords: Counting-independent sets; Graph distance; Edge ideal metric; Hilbert polynomial; Cliques
Fig. 1. Graphs G1 and G2 of Example 1.
Fig. 2. Representation of an α helix, two parallel β strands and two anti-parallel β strands.
Fig. 3. The linear dependency between i5 and d6 for 231 protein pairs.
Fig. 4. On the left, a set of points. In the middle, three points that are neighbours because the shown circle does not contain any other point of the set. On the right, the Voronoi neighbours of this set.

Corresponding author. Tel.: +34 971172969; fax: +34 971173003.